Supplementary information. for the paper:

Transcription

1 Supplementary information for the paper: Screening for protein-protein interactions using Förster resonance energy transfer (FRET) and fluorescence lifetime imaging microscopy (FLIM) Anca Margineanu 1*, Jia Jia Chan 2**, Douglas J. Kelly 1,3**, Sean C. Warren 1,3**, Delphine Flatters 4, Sunil Kumar 1, Matilda Katan 2, Christopher W. Dunsby 1, Paul M.W. French 1* 1 Imperial College London, Dept. Physics, Photonics Lab., Blackett building, Prince Consort Road, London, SW7 2AZ, UK 2 University College London, Institute of Structural and Molecular Biology, Darwin building, Gower St., London, WC1E 6BT, UK 3 Imperial College London, Institute of Chemical Biology, London, SW7 2AZ, London, UK 4 Université Paris Diderot, Sorbonne Paris Cité, Molécules Thérapeutiques in silico, Inserm UMR- S 973, 35 rue Helene Brion, Paris, France *To whom the correspondence should be addressed: Anca.Margineanu@imperial.ac.uk paul.french@imperial.ac.uk **These authors contributed equally to this work 1

2 SARAH domain contact interface/ interaction mode Figure S.1. Heterodimeric models of the SARAH MST1 and SARAH RASSF domains. A) Multiple sequence alignments of the Prosite predicted SARAH domain. The secondary structural elements are indicated as helix 1 (H1) and helix 2 (H2). Non-polar residues that are critical for the hydrophobic framework are marked with asterisks (*). Important residues involved in dimerisation are also highlighted according to their properties: non-polar (yellow), acidic (red) and basic (blue). B) The structure of the MST1 monomer is in purple and the RASSF monomer in cyan. Non-polar (yellow), acidic (red) and basic (blue) side chains of the residues involved in the heterodimeric interface are shown. 2

3 Table S.1 Interface sizes (in Å 2 ) calculated using naccess for the best SARAH heterodimer model of each RASSF with MST1. MST1 Total Side chain Main chain Non-polar Polar with: RASSF RASSF RASSF RASSF RASSF RASSF Double exponential analysis of the plate measured with the Nipkow disk configuration (sectioned) For the double exponential analysis, the donor lifetime in the absence of FRET ( D) is calculated by averaging the values of the EGFP lifetimes obtained in the wells containing the donor only, while the values of the donor in the presence of the acceptor ( DA) are obtained by global analysis after fixing the donor only lifetime. These values are listed in the table S.2. Based on the donor only lifetime D and the lifetime of the donor in the presence of the acceptor DA, the FRET efficiency E FRET can be calculated, from which the donor-acceptor distance r is typically estimated in FRET experiments given that the Förster radius R 0 is known for the specific fluorophores pair: E = 1 DA D = ( R 0 r+r 0 ) 6 Eq. S1 The R 0 estimation is generally done with the assumption that the orientation angles between the donor and acceptor molecules are randomly distributed. This is considered to be true in the case of small dye molecules that can undergo fast rotation on time scales shorter than the fluorescence lifetime. It has been recently demonstrated that in the case of fluorescent proteins this assumption cannot hold because their rotational correlation times (15-20 ns) are much longer that their fluorescence decay time (2-4 ns) (Vogel S.S., Nguyen T.A., van der Meer B.W., Blank P.S. The impact of heterogeneity and dark acceptor states on FRET: Implications for using fluorescent protein donors and acceptors. PLoS ONE 7, e49593, 2012). Thus, the fluorescent proteins are constrained to a given orientation during the FRET measurements, and the orientation factor (depending on the angle between the two dipoles) must be taken into account as well as the donor-acceptor distance when calculating E FRET. 3

4 As it is very difficult to measure experimentally the orientation factor, Vogel et al. (Vogel S.S., van der Meer B.W., Blank P.S. Estimating the distance separating fluorescent protein FRET pairs. Methods 66, , 2013) proposed an empirical relation to determine r in the case of fluorescent proteins, and we have used this reference to obtain the values shown in table S.2. Table S.2. EGFP lifetimes obtained from double exponential analysis, FRET efficiencies E FRET and donoracceptor distances r for the RASSF-SARAH MST1 and RASSF-full length MST1 interactions. SARAH MST1 D (ps) DA (ps) E FRET r (nm) D (ps) DA (ps) E FRET r (nm) RASSF RASSF RASSF RASSF RASSF RASSF Derivation of the equation for K D calculation Dissociation constants K D were calculated for a bi-molecular reaction as described by equation S1, where D is the donor-labelled binding partner, A is the acceptor-labelled binding partner and DA is the complex formed by their association: D free + A free DA Eq. S1 K D is then given by equation S2, which relates the concentrations of the binding partners to the complex. K D = [D free][a free ] [DA] Eq. S2 Using the fluorophore concentration calibration (figure 10B,C in the text of the paper), we can determine the total donor (D total) and acceptor (A total) concentrations, while the FRET fraction obtained from the 4

5 FLIM global analysis provides an estimate of the concentration of the DA complex via the bound fraction of the donor. We can then write: [DA] = D total [DA] = A total Eq. S3 Eq. S4 where is the bound fraction of the acceptor molecules within the complex. From the equality of Eq. S3 and S4, this fraction can be determined: = D total A total Eq. S5 Knowing the bound D and A fractions, we can obtain the free fractions: [D free ] = (1 ) D total Eq. S6 [A free ] = (1 ) A total = (1 D total A total ) A total Eq. S7 Replacing Eq. S6 and S7 in the K D expression, we obtain: K d = [D free][a free ] [DA] = (1 ) D total(1 Dtotal )A A total total = D total c (1 ) (1 D I D )c c A I A I A A Eq. S8 where I D and I A are the initial fluorescence intensity of the donor and the acceptor respectively, which are linearly proportional to the donor and acceptor concentrations via the proportionality constants c D and c A (as shown in figure 10B,C in the text of the paper). Tables of changes in donor fluorescence lifetime Tables S.3-S.5 compare the difference in EGFP lifetime between RASSF proteins interacting with SARAH MST1 and the negative control (MST1 SARAH or mcherry). 5

6 Table S.3. Differences in mean fluorescence lifetimes between RASSF1-10 interacting with SARAH MST1 and the negative control (MST1 SARAH) (same data as presented in figure 4 in the text of the paper). Mean difference in EGFP lifetime (ps) RASSF1 233 RASSF2 244 RASSF3 279 RASSF4 248 RASSF5 314 RASSF6 138 RASSF7 25 RASSF8 6 RASSF9 12 RASSF

7 Table S.4. Mean lifetime differences between RASSF1 mutants interacting with SARAH MST1 and with the negative control (mcherry) (data presented in figure 5 in the text of the paper). Mean difference of EGFP lifetime (ps) RASSF1 240 RASSF1 L301P 2 RASSF1 L305P 5 RASSF1 L308P 17 Table S.5. Mean lifetime differences between RASSF5 mutants interacting with SARAH MST1 and with the negative control (mcherry) (data presented in figure 6 in the text of the paper). Mean difference of EGFP lifetime (ps) RASSF5 294 RASSF5 L224P 201 RASSF5 L228P 109 RASSF5 L231P 165 7

8 Examples of dissociation constants K D for homo- and heterodimerisation of RASSF and MST proteins from literature data The following data have been taken from references 26 and 39-41, as numbered in the text of the paper. NORE1 is the alternative name for RASSF5. The numbers between brackets represent the aminoacid sequence that was used in the study. Point mutations (e.g. L 444 P) are also indicated. Abreviations: NMR = nuclear magnetic resonance, ITC = Isothermal titration calorimetry, SFF = stopped-flow fluorimetry, CD = circular dichroism, CPM = 7-diethylamino-3-(4 maleimidylphenyl)-4-methylcoumarin, FM = fluorescein- 5-maleimide, ecfp = enhanced cyan fluorescent protein, eyfp = enhanced yellow fluorescent protein. Table S.7. K D values for NORE1 (RASSF5) and MST1 homo- and heterodimerisation published in literature. 8

9 Protein1 Protein2 K D (µm) Method Ref. RASSF5 SARAH ( ) ( ) ( ) L 444 P ( ) C 220 S/R 242 C-CPM ( ) C 220 S/R 242 C-CPM ( ) C 220 S/C 414 -CPM ( ) C 220 S/R 242 C-CPM ( ) C 220 S/C 414 -CPM ( ) F 487 C-CPM ( ) E 460 C-CPM ( )-eCFP ( )-eCFP ( ) C 220 S/C 414 -CPM ( )-eCFP ( ) ( ) ( ) ( ) L 444 P ( ) C 220 S/R 242 C-FM ( ) C 220 S/C 414 -FM ( ) C 220 S/C 414 -FM ( ) C 220 S/C 414 -eyfp ( ) C 220 S/C 414 -eyfp ( )-eYFP ( ) E 460 C-FM ( )-eYFP ( )-eYFP ( )-eYFP ( ) L 444 P-eYFP ( ) Low nm range NMR ± 16.4 (at 25 C) ITC ± 20.5 (at 30 C) 34.1 ± 4.6 (at 25 C) ITC FRET in SFF FRET in SFF FRET in SFF FRET in SFF FRET in SFF 39 Low nm range FRET in SFF 39 Low nm range FRET in SFF FRET in SFF FRET in SFF FRET in SFF FRET in SFF ± 0.7 CD 40 9

10 ( ) ( ) ( ) ( ) 0.9 ± 0.1 ITC CD 41 10